New Grad Kernel Engineer for AI Hardware & HPC

Cerebras Systems; Sunnyvale (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

Cerebras Systems is seeking a Kernel Engineer to develop high-performance ML and linear algebra kernels for the Wafer-Scale Engine. You will map workloads to the architecture, optimize low-level code, and collaborate with kernel, compiler, performance, and hardware teams.

The role offers opportunities to work on a massively parallel processor, learn the Cerebras instruction set, memory system, and communication model, and contribute to performance-driven software in a cutting-edge AI hardware

Qualifications

• Bachelor’s degree or higher in a STEM field.
• Proficiency in C++ and Python.
• Solid understanding of processors, memory hierarchies, and data movement.
• Experience with data structures, algorithms, and software fundamentals.
• Experience debugging software through coursework or projects.
• Strong analytical and problem-solving skills.
• Interest in low-level software, parallel computing, or hardware/software co-design.
• Ability to learn unfamiliar systems and collaborate within a technical team.

Responsibilities

• Design and implement ML and linear algebra kernels for the Cerebras Wafer-Scale Engine.
• Develop and debug high-performance kernel routines using low-level programming and Cerebras Software Language.
• Apply parallel programming to map workloads onto Cerebras architecture.
• Use performance data and profiling tools to evaluate kernel behavior and inform design.
• Investigate correctness, performance, and hardware utilization issues.
• Develop unit tests and system-level validation for kernel libraries.
• Collaborate with kernel, compiler, performance, and hardware engineers to improve software.
• Study emerging ML workloads and evolve the kernel library.
• Participate in code reviews and software development processes.
• Understand Cerebras architecture, ISA, memory system, and communication model.

Skills

C++

Python

Computer architecture

Parallel programming

Low-level software

Debugging

Team collaboration

Education

Bachelor's degree in CS/CE/EE/Math

Tools

Cerebras Software Language

Job description

Cerebras Systems is seeking a Kernel Engineer to develop high-performance ML and linear algebra kernels for the Wafer-Scale Engine. You will map workloads to the architecture, optimize low-level code, and collaborate with kernel, compiler, performance, and hardware teams.

The role offers opportunities to work on a massively parallel processor, learn the Cerebras instruction set, memory system, and communication model, and contribute to performance-driven software in a cutting-edge AI hardware